erch includes seme observations mcifie during the tidc.l in- 

 terval covcrod by the -June 15 series. The depth of L.ho 

 mixed Ic-yer on June 15 wrs plotted fpeinst the tidol hour 

 and interpolf:ted vrlues reed off this grrph for the depth 

 Ft the time, in tidf 1 hours, of the observrtions mrne • 

 during this tid£ 1 interval on the 13th f;.nd ]6th. In this 

 innnn'^'r it V/T;s estira?ted thrt the averfge depth of the 

 mixed Iryer was 14 feet greft.r on June 13 rnd 35 feet less 

 on June 16. Accordingly so ss to eliniinnte, so far as 

 possible, long terra or non-tidfl fluctuations, 14 feet 

 were subtracted frora all the June 13 observf. tions and 35 

 feet added to all those nude on June 16. These data were 

 then assembled and averaged for each tidal hour. The means 

 were then smoothed by a running two term averap^. The' 

 results shown in fig. 7 clearly indicate a tidi i periodici- 

 ty. The amplitude is of . course reduced by the smoothing 

 process. Before smoothing but ? f ter correction for the 

 difference bet^reen the cverfge depth on the various days, 

 the amplitude w£ s about 70 feet. 



The possibility thf t the salinity features of the 

 upper 250 meters rs shown in fig. 4, may be subject to 

 vertical tidal movo'^ents of some 20 meters 'ra';';?- 3 r more ■ 

 detailed cnalysis of our data. If the tctuc 1 Si linity 

 distribution even approaches the complexity of the tempera- 

 ture distribution shown by the 'Atlantis' BT traces, then 

 it is obvious thf t such a movement could recount for a 



